Predictors of loss to follow-up (LTFU) among rural residents enrolled in the Kentucky Viral Hepatitis Treatment (KEY TREAT) trial in Appalachian Kentucky.


Author: Jennifer Havens, Takako Schaninger, Michelle Lofwall, April Young, Jeffrey Weiss, Sharon Walsh

Theme: Models of Care Year: 2023

Background: The purpose of KeY Treat is to maximize engagement in HCV treatment by providing screening and incentivized treatment in a low resource rural community impacted by the opioid epidemic.

Methods: Key treat participants are eligible if >18 years old and viremic. Point of care RNA testing is conducted at baseline, along with a questionnaire assessing sociodemographics, substance use/treatment, mental health disorders (generalized anxiety, major depressive and post-traumatic stress – composite variable created due to colinearity), and injection behaviors. Medication (sofosbuvir/velpatasvir) was dispensed at this initial visit for those without drug-drug interaction risk (89%). Multiple logistic regression was utilized to examine factors independently associated with LTFU.

Results: To date, 368 participants have been enrolled; 96.7% initiated therapy and 26.4% were LTFU. The majority of participants were LTFU during the medication phase (85.3%). Almost all (97.2%) enrollees have a history of drug use; 40% reported current use, and of those, 62.1% are people who inject drugs (PWID). Participants with a driver’s license were significantly less likely to be LTFU (adjusted Odds Ratio [aOR]: 0.27, 95% Confidence Interval [CI]: 0.23, 0.63), whereas participants meeting criteria for a mental health disorder were more likely LTFU (aOR: 1.65, 95% CI: 1.01, 2.72). PWID were no more likely than non-PWID to be LTFU. Almost half LTFU had undetectable RNA at their last visit. 

Conclusion: Despite KeY Treat eliminating known barriers, 26.4% of participants were LTFU. Although transportation was provided free of charge, it did require scheduling and impacted LTFU. While KeY Treat also employs a social worker, no direct care was provided for mental health, which may have affected LTFU. While half of those LTFU were no longer transmitting the virus, it is unknown whether they were cured. Lastly, PWID were no more likely LTFU, further demonstrating that they should not be excluded from care. 

Download abstract Download poster